search

arq5x / gemini

a lightweight db framework for exploring genetic variation.
http://gemini.readthedocs.org
MIT License
318 stars 120 forks source link

GATK4 compatibility? #922

Closed SarahBeecroft closed 5 years ago

SarahBeecroft commented 5 years ago

HI,

We've been trying to move to GATK4 from GATK3, but now we are not getting the same variants output when we query the database (even though we're using the same samples). We are pretty stumped, as the process works to completion without errors. I even tried with the joint VCF made from BcBio in case I'd stuffed the 'homemade' pipeline somewhere. It seems that we mainly/only get rubbish/artifact variants now? I also tried without the VQSR step, but the outcome is the same. We use VEP annotation. Any help would be greatly appreciated! Thanks!

brentp commented 5 years ago

Can you share a small ~20 line VCF that, when you load, gives unexpected results? It's hard to know from your description if this is something in gemini or something peculiar about VCF.

SarahBeecroft commented 5 years ago

failed_VCF.txt Ok sure. This is 50 lines but please let me know if you need more. I can also give and example of one that worked. they're both VCF4.2 from what I can see. thanks!!

brentp commented 5 years ago

I mean the first 50 lines after the header (and the header).

SarahBeecroft commented 5 years ago

failed_VCF.txt ah sorry

brentp commented 5 years ago

well, that's easy, that appears to be a gvcf (and every single genotype in there is ./.)

SarahBeecroft commented 5 years ago

That would do it!! thanks very much!